March 23rd, 2016 by Adam Armstrong
Google Cloud Platform Announces New Analytics Tools
Google Cloud Platform has been introducing and working on big data and analytics products and services for some time now. The industry as a whole is seeing the promise in these products and services. Today at Google’s GCP NEXT event, they announced new innovations, products, and services surrounding data management and analytics capabilities. These new products and innovations center around machine learning, big data and analytics, and open source.
Google has introduced a new product family in Cloud Machine Learning. This new product family gives scientists and developers a method of building a new class of intelligent applications while using the same technology that powers Google Now, Google Photos and voice recognition in Google Search. User can build sophisticated, large scale machine learning models in a short amount of time that are portable, fully managed and scalable. The Machine Learning models are built using the open-source TensorFlow machine learning library. Cloud Machine Learning integrates with many of Google Cloud Platform products and customers can use pre-trained Machine Learning models.
Google already offers Cloud DataProc, its managed Apache Hadoop and Apache Spark service, to aid with being more productive when building applications, with faster and better insights, without having to worry about the underlying infrastructure. Google is adding new services and capabilities to its Big Data and Analytics with BigQuery and Google Data Studio 360. Data Studio enables user to unify all analytics workflows into one tool. BigQuery gets several new features including:
- Long Term Storage automatically drops the price of storage by 50% after 90 days.
- Automatic Table Partitions simplifies the way you store and query data by partitioning user tables by date and querying the date ranges users want.
- The new Capacitor storage engine, which accelerates many queries by up to 10x, and Poseidon, a new mechanism that improves data ingest and export speed by 5x.
- Direct query and import of Apache AVRO files, simplifying data interoperability
- Automatic schema detection of JSON and CSV files.
- The new Public Datasets Program to help our community host, share and analyze public datasets.
Google has open sourced some of its machine learning including its machine learning system, TensorFlow. Google states that TensorFlow can now be used with Kubernetes to scale and serve ML models. And these capabilities have been extended helping customers to build powerful Machine Learning models on Cloud Platform.